Seiji Kawamura

A climatology of middle and upper atmosphere radar observations of thermospheric winds

Shigaraki middle and upper atmosphere (MU) radar observations of horizontal thermospheric winds in the magnetic meridian plane over the period September 1986 to September 1996 are reported as climatological averages in the form of time-of-day variations for several combinations of seasonal and solar activity conditions and are compared with winds predicted by the horizontal wind model (HWM) and with winds measured at Saint Santin and Millstone Hill. The dominant feature of the MU wind behavior is its mean diurnal variation of northward flow by day and southward flow by night, with the nighttime wind smoothly approaching and receding from a midnight maximum, while the daytime wind tends to show two peaks, a strong one in the early daylight hours and a weak one in the afternoon-evening. HWM shows the same unimodal nighttime and bimodal daytime behavior, but the HWM pattern is shifted about 2 hours later in time. The amplitude of the diurnal harmonic decreases from 78 m/s at solar minimum to 45 m/s at solar maximum, while HWM shows a corresponding increase from 53 to 62 m/s. The diurnal amplitude is remarkably stable with season but is superposed on a steady wind of 41 m/s southward in summer, 15 m/s northward in winter, and midway between these limits at the equinoxes. HWM shows a symmetric pattern of 30 m/s southward in summer and 30 m/s northward in winter. Ion drag appears to be the main regulator of wind speed, and the seasonal wind patterns have a profound effect on the seasonal behavior of the ionosphere.

Plasma temperature variations in the ionosphere over the middle and upper atmosphere radar

The temperature variations in the ionosphere over the middle and upper atmosphere radar at Shigaraki (34.85°N, 136.10°E, magnetic latitude 25°N) in Japan are studied using the electron and ion temperature (Te and Ti, respectively) data measured by the radar during nearly a full solar cycle (1986–1995). A comprehensive picture of the diurnal, seasonal, and solar activity variations of Te and Ti is presented for the altitude range 200–550 km. The temperatures Te and Ti are found to have similar diurnal and altitude variations and different seasonal and solar activity dependence. With season, while daytime Te is highest in summer and lowest in equinox, daytime Ti is highest in equinox and lowest in summer. With solar activity, while daytime Te decreases, the corresponding Ti increases. The diurnal variation of Te is characterized by morning and evening peaks. The occurrence and strength of these peaks are found to depend on altitude, season, and solar activity. The peaks arise basically from the photoelectron heating of the morning and evening electron gas. However, neutral winds play a dominant role in the appearance of the peaks. A poleward wind, which reduces the electron density to a low value before sunset, is an essential requirement, especially for the evening peak. The mechanisms causing the morning and evening peaks in Te are illustrated through model calculations using the Sheffield University plasmasphere-ionosphere model.

Extraction of solar and thermospheric information from the ionospheric electron density profile

This exploratory study of ionospheric electron density N-e profile assimilation concerns the extraction of information on exospheric temperature T-ex, atmospheric composition [O] and [N-2], meridional thermospheric winds, and solar EUV flux. These parameters are used as input variables to an ionospheric model and adjusted, one, two, or three at a time, to bring the model N-e profile into best agreement with an experimental N-e profile. The important ambiguity/uniqueness issue in the multiple-variable fit, which has yet to be addressed in the literature, is discussed. We found the following: (1) In a single-variable fit, any variable tested may be adjusted successfully to improve the data-model fit, but the validity of the derived variable generally depends on the validity of the values chosen for the fixed parameters. (2) The best fit model profile from a single-variable fit sometimes shows clear differences from the measured profile, indicating that multiple-variable fitting is warranted. The multiple-variable fit often greatly reduces the model-data profile differences and improves the error bars on the derived parameters. (3) Multiple-variable fits often show large variable correlations, indicating that variable determination is often not unique. (4) The meridional wind often shows weak correlation with other variables, suggesting that the meridional wind and any other parameter can be reliably extracted, provided that the other atmospheric variables are known. (5) [N-2] often shows strong correlation with other parameters (except with the wind). The EUV flux and [O] also show large correlation. (6) The T-ex-[O] and T-ex-EUV pairs show medium to high correlation. (7) The three-variable fits nearly always show large ambiguities in the derived parameters.

Simultaneous mesosphere-lower thermosphere and thermospheric F region observations using middle and upper atmosphere radar

Simultaneous MLT (mesosphere-lower thermosphere) and thermospheric F region (upper thermosphere and ionosphere together) observations conducted using the middle and upper atmosphere (MU) radar (35 degrees N, 136 degrees E) in alternate meteor and incoherent scatter modes in October 2000 and March 2001 are presented. The continuous observations, each lasting more than a week, provide simultaneous zonal and meridional wind velocities at MLT altitudes (80-95 km), meridional wind velocity in the upper thermosphere (220-450 km), and electron density and peak height in the ionosphere with a time resolution of 1.5 hours. The data seem to suggest that the upper atmospheric regions could be dynamically coupled through mean winds, tides, and waves. Diurnal (24-hour) and semidiurnal (12-hour) tides and waves of periods 16-20 hours and 35-55 hours coexist at MLT and upper thermosphere altitudes, and the waves become stronger than tides at mesopause (approximate to 88 km) in both October and March. The data in these equinoctial months also show large differences in mean winds, tides, and waves in the MLT region. The amplitudes and phases of the 24-hour and 12-hour tides at MLT altitudes are compared with those predicted by the global scale wave model (GSWM). The model qualitatively predicts the observed growth of the tides with altitude but does not predict the 12-hour tide becoming stronger than the 24-hour tide at altitudes above mesopause in October.

Coordinated airglow observations between IMAP/VISI and a ground-based all-sky imager on concentric gravity wave in the mesopause

We present a study of concentric gravity waves (CGWs) event from the coordinated observation between Ionosphere, Mesosphere, upper Atmosphere, and Plasmasphere mapping (IMAP)/Visible and near-Infrared Spectral Imager (VISI), all-sky camera at Rikubetsu, Multi-functional Transport Satellite (MTSAT), Tropical Rainfall Measuring Mission, and MF radar at Wakkanai combined with Modern-Era Retrospective Analysis for Research and Application data. IMAP/VISI is the first space-based imager that capable of imaging the airglow in the mesosphere and lower thermosphere region in the nadir-looking direction. Therefore, it has a unique ability to observe a great extend of CGWs propagation. Arc-like shaped, part of CGWs pattern was observed around the mesopause (~95 km) in the O2 762 nm airglow emission obtained by IMAP/VISI at 1204 UT on 18 October 2012. Similar patterns were also observed by the all-sky imager at Rikubetsu (43.5°N, 143.8°E) in OI 557.7 nm and OH band airglow emissions from ~1100 to 1200 UT. Horizontal wavelengths of the observed small-scale gravity waves are ~50 km (OH band and OI 557.7 nm) and ~67 km (O2 762 nm). The source is suggested to be a deep convective activity over Honshu Island which likely was an enhanced convective activity related to a typhoon in the south of Japan. The data showed that the CGWs could propagate up to ~1400–1500 km horizontally from the source to the mesopause but not farther away. Using atmospheric temperature profiles obtained by Thermospheric Ionosphere Mesosphere Energetics Dynamics/Sounding of the Atmosphere using Broadband Emission Radiometry, we conclude that this long-distance propagation of the waves could be caused by thermal duct in the middle atmosphere. The arc-like shaped instead of full circle pattern points out that the wind filtering effect is significant for the particular direction of wave propagation.

Tidal waves in the polar lower thermosphere observed using the EISCAT long run data set obtained in September 2005

[1] Characteristics are presented of the lower thermospheric wind from a long run data set obtained by the EISCAT UHF radar at Tromso (69.6°N, 19.2°E) over ~23 days, from September 6 to 29, 2005. The derived semidiurnal amplitude exhibited day-to-day variations (~5-30 ms -1 ) at and above 109 km, while the phase varied little with the day. We have found a mode change of the semidiurnal tide occurring during September 17-22, 2005. Between September 6 and 16, the vertical wavelengths were estimated to be ~58 km and ~76 km for the meridional and zonal components, respectively, while between September 23 and 29, they became less than -24 km. The day to day variability of the diurnal tide was less obvious than that of the semidiurnal tide. The diurnal amplitude of the meridional component increased with height except for 8 days between September 13 and 20, when the diurnal amplitudes were smaller values (<40 ms -1 ) at and above 111 km than those for the other intervals. Furthermore, the shapes of the altitude profiles of the meridional phase differ from those for the other intervals. We have evaluated contributions due to the electric field and the ion drag acceleration and showed that they were not the causes. From the analysis of 22.5 days of wind data, we found about 5-6 day oscillations in the lower thermosphere, probably where there were planetary wave activities in the lower thermosphere.

MU radar observations of H+ ions in the topside ionosphere

We report the detection of H+ ions in the topside ionosphere above the middle and upper atmosphere (MU) radar in Japan, the first such detection by a mesosphere-stratosphere-troposphere radar extended to perform as an incoherent scatter radar. Owing to the very limited signal-to-noise ratios of ∼1% achievable in the topside ionosphere with this radar, long integration times of 45 min and statistical analysis of 10 years of data are employed to yield useful results. This sensitivity restriction limits the top altitude of observation to only ∼400 km during solar-minimum winter but up to ∼650 km during solar-maximum summer. These measurements are possible only at night, when we may assume that the electrons and ions have the same temperature. While the data scatter is large, climatological averages constructed for various solar-activity, seasonal, and magnetic activity cases show less than ∼5% H+ at most under conditions of detect ability, although it is exactly those conditions of undetectability which are expected to correspond to the highest proportions of H+. At solar maximum we see no detectable amounts of H+ up to 650 km altitude, except perhaps for a faint amount in winter. At solar minimum the data are good up to ∼550 km during summer and equinox but to only ∼400 km during winter. The H+ profiles for solar-minimum summer and equinox closely resemble each other and reach ∼5% at 550 km altitude. We see no clearly discernible H+ in winter up to our 400 km altitude limit of detectability; we expect that there is substantial H+ present above 400 km, but the total ion density is so low that we cannot measure it. We see no detectable change in H+ between magnetically quiet and disturbed periods. The H+ proportions seen above the MU radar are much smaller than those seen above Arecibo, which has a similar magnetic latitude but much lower geographic latitude, and are much more comparable to those amounts seen by the higher-latitude radars at Saint Santin and Millstone Hill. Comparisons with models suggest that at these low altitudes, H+ may well be near its charge-exchange-equilibrium value.

Tidal modulations of mesospheric gravity wave kinetic energy observed with MF radar at Poker Flat Research Range, Alaska

The interactions between gravity waves and atmospheric tidal waves have been observationally studied, although the phase relation between them has not been fully understood. In this study, the long-term wind velocity data observed with the Poker Flat MF radar (65∘N, 147∘W) were analyzed for the period of 1999–2008 to show local time dependence and seasonal climatologies of the 12 h and 24 h components in the mesospheric winds and their modulations of gravity wave kinetic energy. We made climatological 1 day composite plots of the kinetic energy of gravity waves for wave periods of 1–4 h and harmonic components of horizontal wind for each month. The results show that the kinetic energy of gravity waves peaks twice at 3–6 LT and 18–21 LT, which tend to coincide with the transition of the 12 h component of zonal wind from westward to eastward flow. On the other hand, a 2 month case study revealed that the gravity wave kinetic energy and the 12 h components of zonal wind appear to keep their phase difference constant (like a “phase locked”) for more than 10 days. Events of this kind are also found in other years. To examine whether this relation can be explained by interaction between the 12 h component of zonal wind and gravity waves, we applied a gravity wave drag model to the background state defined as the sum of observed monthly mean and harmonic components of zonal wind. It is suggested that the orographic gravity wave drag has a 12 h periodicity and that the time of the drag enhancement changes in time following change in the phase of harmonic components of winds.

Precipitation measurement using a dual Ka-band radar system for GPM/DPR algorithm development

A dual Ka-band radar system was developed by the Japan Aerospace Exploration Agency for GPM/DPR algorithm development. The system consists of two identical Ka-band radars. Both the equivalent radar reflectivity factor (Ze) and specific attenuation (k) can be estimated at each range of the path. The estimated k-Ze relations of rain, snow, and the melting layer can be used for the GPM/DPR algorithm development. Those parameters of snow at Ka band are important. Observations using the dual Ka radar system were performed in Okinawa Island, and Nagaoka, Niigata, Japan. In Okinawa Island, rain observation was conducted for confirming the performance of the measurement, and reasonable k-Ze relations of rain were obtained. In Nagaoka, snow observation was conducted. Different k-Ze relations were obtained. This was attributed to the attenuation difference between wet and dry snow.

Water vapor estimation using digital terrestrial broadcasting waves: WATER VAPOR ESTIMATION USING DTB WAVES

近年都市部で頻発する局地的大雨(通称ゲリラ豪雨) などの時空間スケールの小さな気象現象は、孤立した 積乱雲の急激な生成・発達により引き起こされる。従 来のレーダー観測ではとらえることが困難だったこの ような現象が、フェーズドアレイ気象レーダー (PAWR)の登場により可視化できるようになってき た。PAWR は半径 60 km の範囲の雨を 30 秒ごとに 三次元観測することができる。さらに、2018 年 3 月 から埼玉大学で観測が開始されたマルチパラメータ・ フェーズドアレイ気象レーダー(MP-PAWR)は、観 測の高速性を保ちつつ偏波の情報を使ってより定量的 な降雨観測を実現している。このようなフェーズドア レイ気象レーダーを用いることで、上空で急発達した 降水粒子が落下に要する 5 ~ 10 分後には地上のどの あたりにどの程度の降水をもたらすか、といった短時 間の予測が可能になってきた。 しかし、防災・減災のための対応が可能となるよう な長いリードタイムを取った予測(20 分~数時間先の 予測)の精度はいまだ十分とは言えない。このような 長い予測の精度向上には数値予報モデルが重要であり、 その予測精度向上にはモデルそのものの改良に加えて、 より多くの観測データを取り込む(データ同化する) ことが必要となってくる。その中でも特に近年期待さ れているのが水蒸気量のデータ同化である。水蒸気は 雨の元となる気体としての水であり、この水蒸気の動 きを早い段階から連続して監視することで、より精度 の高い降雨予測につながると期待されている。本稿で は、NICT が開発し、現在首都圏を中心に実証実験を 実施している地上デジタル放送波(地デジ放送波)を 用いた水蒸気量推定技術を紹介する。